LETTER TO THE EDITOR Do specialty anticoagulation clinics really outperform primary care at INR management? Scott R. Garrison • G. Michael Allan Published online: 19 July 2014 Ó Springer Science+Business Media New York 2014 To the Editor, Time in therapeutic INR range (TTR) is the most accepted measure of the quality of oral anticoagulation (warfarin) management and it is well established that TTR varies substantially across geographic regions and across clinical settings [1]. In particular, it has been clearly demonstrated that TTR in specialty anticoagulation clinics is higher than TTR in primary care settings. Because specialty clinics typically utilize an algorithm for warfarin dose adjustments there is speculation that such algorithms are a major reason for the clinic’s superior TTR. Indeed, this journal just published an RCT by Nieuwlaat and colleagues evaluating whether the same algorithm used in the RE-LY trial, in the hands of primary care providers, would improve TTR [2]. Despite good concordance with the algorithm however, algorithm-guided warfarin man- agement provided no improvement in TTR over usual primary care decision-making. Should this surprise us? The 2012 ACCP Guidelines for Antithrombotic Therapy and Prevention of Thrombosis identified 3 RCTs comparing ‘‘usual care’’ warfarin dosing by primary care physicians with algorithmic warfarin dosing performed by a variety of other anticoagulation providers. One of these RCTs compared primary care management to algorithmic management by a pharmacist [3]. Another compared primary care management to that of an anticoagulation service [4]. Both studies reported no difference in TTR. In the third RCT patients presenting to a specialty anticoagulation clinic were randomized to have their war- farin managed either by the clinic, or by their own family physician [5]. A statistically significant difference in favor of the specialty clinic was found (82 vs. 76 %) but this difference was less than the clinically important difference the authors had predefined. What is more striking about the reporting from this trial is the indication for anticoagulation of patients attending this Canadian anticoagulation clinic, the majority of whom had venous thromboembolism (64 % venous thromboembolism, 18 % atrial fibrillation, 11 % mechanical heart valve, 7 % other). If this population is typical of specialty clinics then primary care and specialty clinic patients are very different. Primary care physicians can be assumed to manage patients whose indications for anticoagulation distribute as per the population norm. From a Swedish national anti- coagulation registry we might expect 64 % to have atrial fibrillation, 13 % to have valvular heart disease and only 19 % to have venous thromboembolism [6]. In Nieuwlaat, which excluded valvular heart disease, 85 % of partici- pating Canadian primary care warfarin users had atrial fibrillation. The pronounced difference in indication for anticoagu- lation in these two clinical settings brings us to question whether patient characteristics with more influence on TTR might also distribute unequally. Could highly motivated (i.e. more compliant) patients be more likely to seek spe- cialty care? Could frail elderly with mild cognitive impairment, multiple comorbities, greater polypharmacy, more intercurrent illnesses (and hence more frequent pre- scription changes) be more likely to be managed by their primary care provider? In our experience the answer is yes. If such patients do not distribute equally between specialty and primary care settings, is it valid to compare TTR across S. R. Garrison (&) Á G. M. Allan Department of Family Medicine, University of Alberta, 8215-112 Street NW, Room 1706 College Plaza, Edmonton, AB T6G 2C8, Canada e-mail: scott.garrison@ualberta.ca 123 J Thromb Thrombolysis (2014) 38:420–421 DOI 10.1007/s11239-014-1113-2